Before we dive into our economic analysis, we need to set up our analytical environment. This involves importing the necessary Python libraries that will serve as our tools for data manipulation and visualization.
We’ll be using:
Pandas: A powerful library for data manipulation and analysis
Matplotlib: A comprehensive library for creating static, animated, and interactive visualizations
These libraries will allow us to efficiently handle large datasets, perform complex calculations, and create insightful visualizations of economic trends and relationships.
Let’s start by importing these libraries and loading our economic dataset:
import pandas as pdimport matplotlib.pyplot as plt# Load the economic data from our CSV filedf = pd.read_csv('data.csv')# Display the first few rows of the dataset to get an initial overviewdisplay(df.head())# Show the overall shape of our datasetprint("Dataset shape:", df.shape)
Country Name
Country Code
Series Name
Series Code
1960 [YR1960]
1961 [YR1961]
1962 [YR1962]
1963 [YR1963]
1964 [YR1964]
1965 [YR1965]
...
2014 [YR2014]
2015 [YR2015]
2016 [YR2016]
2017 [YR2017]
2018 [YR2018]
2019 [YR2019]
2020 [YR2020]
2021 [YR2021]
2022 [YR2022]
2023 [YR2023]
0
Thailand
THA
GDP (current US$)
NY.GDP.MKTP.CD
2760750861
3034037811
3308912797
3540403457
3889129942
4388937649
...
4.073390e+11
4.012960e+11
4.133660e+11
4.563570e+11
5.067540e+11
5.439770e+11
5.00462E+11
5.06256E+11
4.95645E+11
5.14945E+11
1
Thailand
THA
Population, total
SP.POP.TOTL
26596584
27399963
28242174
29114124
30013573
30940270
...
6.996094e+07
7.029440e+07
7.060704e+07
7.089820e+07
7.112780e+07
7.130776e+07
71475664
71601103
71697030
71801279
2
Thailand
THA
Unemployment, total (% of total labor force) (...
SL.UEM.TOTL.ZS
..
..
..
..
..
..
...
5.760000e-01
5.970000e-01
6.880000e-01
8.300000e-01
7.660000e-01
7.160000e-01
1.099
1.215
0.94
0.913
3
Thailand
THA
Inflation, consumer prices (annual %)
FP.CPI.TOTL.ZG
-0.765864333
7.386990077
3.696098563
..
-0.792079208
0.166333998
...
1.895142e+00
-9.004250e-01
1.881497e-01
6.656319e-01
1.063898e+00
7.067286e-01
-0.845937147
1.230395413
6.077412284
1.228026235
4
Thailand
THA
Consumer price index (2010 = 100)
FP.CPI.TOTL
10.21121022
10.96551131
11.37080742
11.37080742
11.28074161
11.29950532
...
1.113459e+02
1.103433e+02
1.105509e+02
1.112868e+02
1.124708e+02
1.132656e+02
112.3074938
113.68932
120.5986887
122.0796723
5 rows × 68 columns
Dataset shape: (80, 68)
By executing this code, we’ll get our first look at the economic data we’ll be working with. The head() function will show us the first few rows of the dataset, giving us an idea of its structure and the types of economic indicators it contains. The shape attribute will tell us how many rows and columns are in our dataset, indicating the breadth of our economic data.
['GDP (current US$)' 'Population, total'
'Unemployment, total (% of total labor force) (modeled ILO estimate)'
'Inflation, consumer prices (annual %)'
'Consumer price index (2010 = 100)']
Restructuring the Data for Time Series Analysis
Now that we’ve loaded our data, we need to restructure it into a format that’s more suitable for time series analysis. Our current dataset likely has years as separate columns, which isn’t ideal for analyzing trends over time or comparing different economic indicators. We’ll also take this opportunity to simplify our data by renaming some columns and dropping unnecessary information.
Here’s what we’re going to do:
Create a dictionary to store these renamed mappings for future use
Rename the economic indicators in the Series Name column for easier reference
Drop the Country Code and Series Code column as we won’t be using it in our analysis
Convert our wide-format data (where each year is a separate column) into a long format
Create a single Year column and a corresponding Value column
Let’s start by creating a dictionary for our renamed series:
new_series_names = ['GDP', 'Population', 'Unemployment', 'Inflation', 'CPI']series_names =dict(zip(old_series_names, new_series_names))# Display our mapping dictionaryprint("Series name mappings:")for old_name, new_name in series_names.items():print(f"{old_name} -> {new_name}")# Rename the series using our mapping dictionarydf['Series Name'] = df['Series Name'].replace(series_names)
Series name mappings:
GDP (current US$) -> GDP
Population, total -> Population
Unemployment, total (% of total labor force) (modeled ILO estimate) -> Unemployment
Inflation, consumer prices (annual %) -> Inflation
Consumer price index (2010 = 100) -> CPI
# Melt the dataframe to convert years from columns to rowsdf_melted = df.melt(id_vars=['Country Name', 'Series Name'], var_name='Year', value_name='Value')print(df_melted.head())
Country Name Series Name Year Value
0 Thailand GDP 1960 [YR1960] 2760750861
1 Thailand Population 1960 [YR1960] 26596584
2 Thailand Unemployment 1960 [YR1960] ..
3 Thailand Inflation 1960 [YR1960] -0.765864333
4 Thailand CPI 1960 [YR1960] 10.21121022
Check data type of each columns.
print(df_melted.dtypes)
Country Name object
Series Name object
Year object
Value object
dtype: object
Clean up the DataFrame.
# Clean up the 'Year' column by removing any non-year text and convert to integerdf_melted['Year'] = df_melted['Year'].str[:4].astype(int)# Convert 'Value' to numeric, handling any non-numeric valuesdf_melted['Value'] = pd.to_numeric(df_melted['Value'], errors='coerce')print(df_melted.head())
Country Name Series Name Year Value
0 Thailand GDP 1960 2.760751e+09
1 Thailand Population 1960 2.659658e+07
2 Thailand Unemployment 1960 NaN
3 Thailand Inflation 1960 -7.658643e-01
4 Thailand CPI 1960 1.021121e+01
print(df_melted.dtypes)
Country Name object
Series Name object
Year int64
Value float64
dtype: object
This transformation gives us a more flexible and streamlined dataset. Each row now represents a specific economic indicator for a particular country in a given year, with the indicator names simplified for easier reference.
The benefits of this restructured data include:
Easier filtering and grouping by year, country, or economic indicator
Simplified column names for more intuitive data manipulation
A consistent format for time series analysis across different economic indicators
Removed unnecessary data to focus on the most relevant information
This format will make it much easier to perform time series analysis, compare different countries, or analyze trends across various economic indicators in our subsequent analyses.
Exploratory Data Analysis
Now that we have our data in a suitable format, let’s explore it to get a better understanding of our economic indicators across different countries and years.
# Get basic statistics for GDP indicator of Thailandcountry ='Thailand'indicator ='GDP'print(df_melted[(df_melted['Country Name'] == country) & (df_melted['Series Name'] == indicator)]['Value'].describe())
count 6.400000e+01
mean 1.592162e+11
std 1.722203e+11
min 2.760751e+09
25% 1.645960e+10
50% 1.048439e+11
75% 2.676350e+11
max 5.439770e+11
Name: Value, dtype: float64
We may loop through all avaliable indicators to get basic statistics for each economic indicator.
for indicator in df_melted['Series Name'].unique():print('\n')print(indicator)print(df_melted[(df_melted['Country Name'] == country) & (df_melted['Series Name'] == indicator)]['Value'].describe())
GDP
count 6.400000e+01
mean 1.592162e+11
std 1.722203e+11
min 2.760751e+09
25% 1.645960e+10
50% 1.048439e+11
75% 2.676350e+11
max 5.439770e+11
Name: Value, dtype: float64
Population
count 6.400000e+01
mean 5.376486e+07
std 1.447177e+07
min 2.659658e+07
25% 4.163871e+07
50% 5.651944e+07
75% 6.695213e+07
max 7.180128e+07
Name: Value, dtype: float64
Unemployment
count 33.000000
mean 1.301909
std 0.743189
min 0.249000
25% 0.766000
50% 1.180000
75% 1.490000
max 3.404000
Name: Value, dtype: float64
Inflation
count 63.000000
mean 4.096830
std 4.560833
min -0.900425
25% 1.229211
50% 3.312192
75% 5.411977
max 24.313560
Name: Value, dtype: float64
CPI
count 64.000000
mean 58.696073
std 37.776153
min 10.211210
25% 20.990245
50% 53.688850
75% 93.675617
max 122.079672
Name: Value, dtype: float64
Check for missing values
print("Missing values in each column:")print(df_melted.isnull().sum())
Missing values in each column:
Country Name 0
Series Name 0
Year 0
Value 974
dtype: int64
Get the range of years in our dataset
print(f"Year range: {df_melted['Year'].min()} to {df_melted['Year'].max()}")
Year range: 1960 to 2023
We may need to drop those missing value in Value column.
print("Melted dataset shape (before drop NaN):", df_melted.shape)df_melted.dropna(inplace =True)print("\nAfter drop NaN")print("Missing values in each column:")print(df_melted.isnull().sum())print("Melted dataset shape:", df_melted.shape)
Melted dataset shape (before drop NaN): (5120, 4)
After drop NaN
Missing values in each column:
Country Name 0
Series Name 0
Year 0
Value 0
dtype: int64
Melted dataset shape: (4146, 4)
Analyzing Economic Growth
Let’s analyze the GDP growth rate for a specific country over time.
.pct_change(): Computes the fractional change from the immediately previous row by default. This is useful in comparing the fraction of change in a time series of elements.
df_melted = df_melted.dropna(axis='index') # Drop rows with NaNdef calculate_gdp_growth(country): gdp_data = df_melted[(df_melted['Series Name'] =='GDP') &# filter data by country (df_melted['Country Name'] == country)] gdp_data = gdp_data.sort_values('Year') # sort data gdp_data['GDP_Growth'] = gdp_data['Value'].pct_change() *100# calculatereturn gdp_datacountry ='China'gdp_growth_data = calculate_gdp_growth(country)print(gdp_growth_data.head())
Country Name Series Name Year Value GDP_Growth
10 China GDP 1960 5.971625e+10 NaN
90 China GDP 1961 5.005669e+10 -16.175770
170 China GDP 1962 4.720919e+10 -5.688550
250 China GDP 1963 5.070661e+10 7.408363
330 China GDP 1964 5.970813e+10 17.752143
plt.figure(figsize=(12, 5))plt.plot(gdp_growth_data['Year'], gdp_growth_data['GDP_Growth'])plt.scatter(gdp_growth_data['Year'], gdp_growth_data['GDP_Growth'])plt.title(f"{country}'s GDP Growth Rate Over Time")plt.xlabel("Year")plt.ylabel("GDP Growth Rate (%)")plt.grid(True)plt.show()
print(f"Average GDP Growth Rate for {country}: \{gdp_growth_data['GDP_Growth'].mean():.2f}%")print(f"Highest GDP Growth Rate: {gdp_growth_data['GDP_Growth'].max():.2f}% in \{gdp_growth_data.loc[gdp_growth_data['GDP_Growth'].idxmax(), 'Year']}")print(f"Lowest GDP Growth Rate: {gdp_growth_data['GDP_Growth'].min():.2f}% in \{gdp_growth_data.loc[gdp_growth_data['GDP_Growth'].idxmin(), 'Year']}")
Average GDP Growth Rate for China: 9.92%
Highest GDP Growth Rate: 30.15% in 1995
Lowest GDP Growth Rate: -16.18% in 1961
This analysis allows us to visualize a country’s economic growth trajectory and identify periods of rapid growth or recession.
Comparing Economic Development Across Countries
Next, let’s compare the GDP per capita across several countries.
Country Name Series Name_x Year Value_x Series Name_y Value_y \
0 Thailand GDP 1960 2.760751e+09 Population 26596584.0
1 China GDP 1960 5.971625e+10 Population 667070000.0
2 India GDP 1960 3.702988e+10 Population 445954579.0
3 Japan GDP 1960 4.741924e+10 Population 93216000.0
4 Thailand GDP 1961 3.034038e+09 Population 27399963.0
GDP_per_capita
0 103.800957
1 89.520214
2 83.035102
3 508.702779
4 110.731456
plt.figure(figsize=(12, 6))for country in countries: country_data = gdp_per_capita_data[gdp_per_capita_data['Country Name'] == country] plt.plot(country_data['Year'], country_data['GDP_per_capita'], marker='o', label=country)plt.title("GDP per Capita Comparison")plt.xlabel("Year")plt.ylabel("GDP per Capita (current US$)")plt.legend()plt.grid(True)plt.show()
# Print the most recent GDP per capita for each countrylatest_year = gdp_per_capita_data['Year'].max()latest_data = gdp_per_capita_data[gdp_per_capita_data['Year'] == latest_year]for country in countries: country_gdp_per_capita =\ latest_data[latest_data['Country Name'] == country]['GDP_per_capita'].values[0]print(f"{country}'s GDP per capita in {latest_year}: ${country_gdp_per_capita:,.2f}")
China's GDP per capita in 2023: $12,614.07
India's GDP per capita in 2023: $2,484.85
Japan's GDP per capita in 2023: $33,834.43
Thailand's GDP per capita in 2023: $7,171.81
This comparison helps us understand the relative economic development of different countries over time.
def compare_countries_bar(indicator, year, top_n=10): data = df_melted[(df_melted['Series Name'] == indicator) & (df_melted['Year'] == year)] data = data.sort_values('Value', ascending =True).tail(top_n)# Convert GDP to billions of dollarsif indicator =='GDP': data.loc[:, 'Value'] = data['Value'] /1e9# Convert to billions plt.figure(figsize=(12, 8)) plt.barh(data['Country Name'], data['Value']) plt.title(f"Top {top_n} Countries by {indicator} in {year}") plt.xlabel("GDP (billion USD)"if indicator =='GDP'else indicator) plt.ylabel("Country") plt.tight_layout() plt.show()compare_countries_bar('GDP', 2020)
Inflation Analysis
Let’s analyze inflation trends for a specific country.
Year GDP_Growth Country Name Series Name Value
0 1991 9.874601 Indonesia Unemployment 2.617
1 1992 9.779458 Indonesia Unemployment 2.734
2 1993 23.416935 Indonesia Unemployment 2.782
3 1994 11.952002 Indonesia Unemployment 4.366
4 1995 14.268593 Indonesia Unemployment 4.611
plt.figure(figsize=(6, 6))plt.scatter(relationship_data['GDP_Growth'], relationship_data['Value'])plt.title(f"GDP Growth vs Unemployment Rate in {country}")plt.xlabel("GDP Growth Rate (%)")plt.ylabel("Unemployment Rate (%)")plt.grid(True)plt.show()correlation = relationship_data['GDP_Growth'].corr(relationship_data['Value'])print(f"Correlation between GDP Growth and Unemployment Rate in {country}: {correlation:.2f}")
Correlation between GDP Growth and Unemployment Rate in Indonesia: 0.23
This analysis helps us understand the relationship between economic growth and labor market conditions, potentially revealing Okun’s Law in action.
These steps provide a comprehensive analysis of various economic indicators, allowing students to derive insights about economic growth, development, monetary policy, and labor markets across different countries and time periods.
Creating an Economic Dashboard
Finally, let’s create a comprehensive economic dashboard for a country, bringing together multiple economic indicators.
------- Unpivot -------
Country Name Series Name Year Value
0 Thailand GDP 1960 2.760751e+09
3 Thailand Inflation 1960 -7.658643e-01
80 Thailand GDP 1961 3.034038e+09
83 Thailand Inflation 1961 7.386990e+00
160 Thailand GDP 1962 3.308913e+09
------- Pivot -------
Series Name GDP Inflation Unemployment
Year
1960 2.760751e+09 -0.765864 NaN
1961 3.034038e+09 7.386990 NaN
1962 3.308913e+09 3.696099 NaN
1963 3.540403e+09 NaN NaN
1964 3.889130e+09 -0.792079 NaN
